Episodic Memory
Virtual Maze Navigation
Also called: virtual navigation task, virtual water maze task, spatial navigation task, virtual maze learning task
Measure spatial learning and memory: participants learn where a star is hidden in a maze seen from above, then reach it from corners never used before.
5-10 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Preview any test free after you sign up.
- Papers, 2021 to 2025
- 199
- Evidence grade
- C
- Registered clinical studies
- 14
What it measures
Measures spatial learning and memory with a hidden star in a maze seen from above. Participants walk through a maze our code grows, seeing only along the paths from where they stand, and learn where the star is over several starts from the sides. The main score is how directly they reach it from corners never used before.
Constructs
- Spatial learning
- Spatial memory
- Wayfinding
Research uses
- Spatial memory and navigation research
- Aging and memory studies
- Hippocampal function research
- Within-study group comparison
How the paradigm works
Virtual maze navigation measures spatial learning and memory with the logic of the water maze (Morris, 1984): a goal hidden in one place is learned over trials that start from different sides, then tested from a start never used before. It follows the computer form of Commins and colleagues (2020), in which a platform hidden in the middle of one quadrant is learned over trials of 60 seconds from the north, south, east and west, and a probe starts from a new place. Here the arena is a maze seen from above, grown by our code from the session's seed, and the participant moves a marker one cell at a time with the arrow keys or the arrow buttons. Only the paths in sight from where the marker stands are shown and the rest of the maze is dark, so its layout has to be learned. A star hidden in the middle of one quadrant stays there for the whole test and shows only when its cell is entered.
More about the paradigm
The standard session has a 7 by 7 maze with loops, 8 learning trials (two rounds that start once from the middle of each side, in a seeded order) and 3 probe trials that start from the corners outside the star's quadrant. A trial ends when the star is found or after 60 seconds; when a learning trial reaches its limit, the star is shown for 2 seconds. Two practice trials in a small maze come first. Every move is kept with its time and whether it came from a key or a button.
The main score is probe path efficiency: on each probe trial, the shortest route to the star divided by the moves taken to reach it, so 1 is the shortest route and 0 means the star was not reached in time, averaged over the probe trials. The same ratio over the learning trials after the first, the moves of the first search, the trials in which the star was found, presses toward a wall, the time between moves and the total time are also reported. Presses under 100 ms after a maze appears are counted as anticipations and do not move the marker. Scores support research and do not diagnose.
Scores
- Probe Path Efficiencyscore
- On each probe trial, the shortest route to the star divided by the moves taken to reach it (0 when it was not reached), averaged over the probe trials.
- Higher is better; 1 is the shortest route
- Probe Stars Foundcount
- Probe trials in which the star was reached before the time limit.
- Higher is better
- Learning Path Efficiencyscore
- The same ratio averaged over the learning trials after the first, in which the star's place was still unknown.
- Higher is better
- Learning Stars Foundcount
- Learning trials in which the star was reached before the time limit.
- Higher is better; none found is flagged for review
- First Search Movesmoves
- Moves on the first learning trial, the first search for the star.
- Informational
- Wall Bumpscount
- Presses toward a wall, which do not move the marker, over all test trials.
- Informational
- Median Time per Movems
- Median time from one move to the next.
- Informational; depends on keys or buttons
- Total Timeseconds
- Time from each trial's onset to the star, summed; a trial in which it was not found counts its time limit.
- Informational
- Anticipationscount
- Presses under 100 ms after a maze appeared, which did not move the marker.
- Lower is better
- Planned Trialstrials
- Learning and probe trials the session planned to show.
- Informational
- Total Trialstrials
- Learning and probe trials shown in the scored run.
- Fewer than planned is flagged for review
Use and evidence
199 papers from 2021 to 2025 name this task in their title or abstract, up 39% compared with 2016 to 2020. That ranks it 122 of 395 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 42 |
| 2024 | 34 |
| 2023 | 58 |
| 2022 | 36 |
| 2021 | 29 |
| 2020 | 37 |
| 2019 | 31 |
| 2018 | 26 |
| 2017 | 24 |
| 2016 | 25 |
| 2015 | 26 |
| 2014 | 18 |
| 2013 | 20 |
| 2012 | 16 |
| 2011 | 17 |
| 2010 | 20 |
| 2009 | 17 |
| 2008 | 15 |
| 2007 | 15 |
| 2006 | 6 |
| 2005 | 3 |
| 2004 | 5 |
| 2003 | 7 |
| 2002 | 10 |
| 2001 | 5 |
| 2000 | 1 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 3
- Papers on reliability or validity
- 25
- Registered clinical studies using it as an outcome
- 14
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Maze
- 7 by 7 cells (standard; or 9 by 9), grown from the session's seed by the growing-tree method and then braided: each dead end is opened into a neighbor with probability 0.5, so loops give more than one route.
- View
- Seen from above. Only the cells along open paths from the marker, up to a wall, are lit; the rest of the maze is dark apart from its outer edge. Nothing marks the path walked.
- Star
- Hidden in the middle of one quadrant, drawn from the seed, and in the same place for the whole test. It shows only when its cell is entered.
- Trials
- 8 learning trials (standard; 4 or 12), each round starting once from the middle of each side in a seeded order, then 3 probe trials from the corners outside the star's quadrant. Two practice trials in a 5 by 5 maze come first, the same for everyone.
- Moves
- Arrow keys or on-screen arrow buttons, one cell per press. A held key moves once, a press toward a wall is counted and does not move, and presses under 100 ms after a maze appears are counted as anticipations.
- Time
- 60 seconds per trial (standard; 30 to 120), with 500 ms of blank screen before each trial. A found star stays for 800 ms; at the limit of a learning trial the star is shown for 2,000 ms, and a probe trial ends without showing it.
- Data
- Every move with its time and whether it came from a key or a button, from which the scorer replays each trial on the saved maze.
Water maze studies remove the goal on the probe trial and score the time spent near its place. Here the star stays on the probe trials and the score is how directly it is reached. The source shows an open arena in first person; here a maze is seen from above with only the paths in sight lit.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Scored rows: virtual_maze_navigation_trial from the scored run, one per trial. A trial has no chance level, so accuracy is not judged against chance. A trial in which the marker never moved before the time limit is a missed response: more than max(3, 25%) invalidate and any is flagged. A session in which the star was never found while learning, or a short run, is flagged for review.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. The probe trials test memory for a place the participant has already found; if the star was never found while learning, they measure search instead. A trial with no move at all is not an attempt and signals non-engagement. These limits are ConductScience choices: the sources set none.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Learning trials | 8 (standard) | 4 (one round), 8 (standard), 12 (three rounds) |
| Time per trial (seconds) | 60 | 30 to 120 |
| Maze size | 7 by 7 (standard) | 7 by 7 (standard), 9 by 9 |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Virtual Maze Navigation task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Morris, 1984; computer form after Commins et al., 2020), with the standard settings (learning trials: 8; time per trial (seconds): 60; maze size: 7 by 7). The main score was probe path efficiency (on each probe trial, the shortest route to the star divided by the moves taken to reach it (0 when it was not reached), averaged over the probe trials).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Morris, R. (1984). Developments of a water-maze procedure for studying spatial learning in the rat. Journal of Neuroscience Methods, 11(1), 47-60. https://doi.org/10.1016/0165-0270(84)90007-4The paradigm: a goal hidden in one place and learned from different starts.
- Commins, S., Duffin, J., Chaves, K., Leahy, D., Corcoran, K., Caffrey, M., Keenan, L., Finan, D., & Thornberry, C. (2020). NavWell: A simplified virtual-reality platform for spatial navigation and memory experiments. Behavior Research Methods, 52(3), 1189-1207. https://doi.org/10.3758/s13428-019-01310-5The computer form: a goal in the middle of one quadrant, 60-second trials from four starts, a probe from a new start and the arrow keys.
Source papers
- Commins S, Duffin J, Chaves K, Leahy D, Corcoran K, Caffrey M, Keenan L, Finan D, Thornberry C (2020). NavWell: A simplified virtual-reality platform for spatial navigation and memory experiments. Behavior Research Methods, 52(3), 1189-1207.Paradigm source, Young adults in the source
Questions
How long does the Virtual Maze Navigation take?
About 5 to 10 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Probe Path Efficiency, Probe Stars Found, Learning Path Efficiency, Learning Stars Found, First Search Moves, Wall Bumps, Median Time per Move, Total Time, Anticipations, Planned Trials and Total Trials. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Virtual Maze Navigation used in research?
199 papers from 2021 to 2025 name it in their title or abstract, up 39% compared with 2016 to 2020. That ranks it 122 of 395 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.14.0, counted 31 October 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Virtual Maze Navigation have?
In the published research, 3 papers report its test-retest reliability and 14 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Which settings can a researcher change?
Learning trials (4 (one round), 8 (standard), 12 (three rounds)), time per trial (seconds) (30 to 120) and maze size (7 by 7 (standard), 9 by 9). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.